The goal of this prospective, longitudinal developmental study is to determine the effects of prenatal crack use on the medical and neurodevelopmental outcomes of infants throughout the first 3 years of life, in a rural, health department population which uses crack, the inexpensive, highly-addictive, potent form of cocaine, and no other illicit drugs. Prenatal crack abuse in this population is of added interest because of its potential long-term impact on public assistance programs. Specific aims include: (1) evaluation of the separate and combined effects on the developing child of crack use, maternal and neonatal heath status, infant neurobehavioral status and temperament, maternal factors, and the caregiving environment; (2) determination of dose response effects of prenatal crack use, separate from other infant, maternal and caregiving risk factors, on medical/neurologic status and behavioral/developmental outcome; (3) determination of the relationship of cocaine metabolites in infant urine specimens to early neurobehavioral outcome and their relationship to sequelae. This study will provide new data about populations never before reported on: (1) women receiving public heath care who do not have ready access to drug rehabilitation; (2) women who use crack almost exclusively as their chosen form of cocaine and use no other illegal drugs; (3) a rural, Southern population. Pilot studies estimate at least 8% of the 3,000 births/year at our hospital are crack exposed. 150 mothers admitting use and 150 matched controls will be enrolled to answer research hypotheses. Maternal urine will be drug-screened at the first prenatal visit and psychosocial/drug histories taken at each trimester and followup visit. At delivery, fetal and neonatal complications will be recorded and parent report inventories completed. Infant measures will include urine drug screen, cranial ultrasound, cry recordings, medical/neurologic, and developmental evaluations. At 1, 12, and 24 months, maternal characteristics and the caregiving environment will be assessed during home visits. At 6, 18, and 36 months medical/neurologic, language, developmental and behavioral assessments will be made. All patients will be referred as needed. Using transactional analyses, we will test the hypotheses that: (1) Prenatal crack use, in a dose-dependent fashion, will be related to more fetal/neonatal complications, family dysfunction, poorer maternal caregiving, and poorer development and (2) children with the most crack exposure, the most medical complications, poorest early neurobehavioral outcome, and the most difficult temperament who have mothers with the most psychosocial problems, least adequate caregiving abilities, and the most dysfunctional families are the most likely to demonstrate less adaptive behavior and less optimal developmental outcome. The uniqueness of this study includes access to a rural population of crack users; the ability to collect extensive prenatal drug histories; use of state of the art developmental and caregiving assessment measures; use of transactional analysis for the evaluation of the effects of crack use on developmental outcome; and dose response effects of crack use on short and long-term developmental outcome.